Amazon Uses Viewer Data in TV

Soliciting viewer feedback during the development process makes a lot of sense

Even robots can spin a good yarn.

When Amazon hatched an inhouse production unit in 2010 with the intent to create its own original programming, there was cause for skepticism. True to form, the retail giant went in with a HAL 9000 approach to creativity that seemed liked Silicon Valley hubris run amok.

Seeking to put a twist on the typical close-to-the-vest development process the entertainment business has made standard practice, Amazon has given its massive user base free access to samples of content including full-length TV pilots that Amazon Studios is considering greenlighting. The feedback would then inform the decision as to what content to choose.

That’s not how it’s typically done in Hollywood, of course. In this town, creative choices are more governed by hunches and intuition than calculation and algorithms. But now the gut is giving way to the gigabyte.

It’s a strategy perfectly in keeping with Amazon, which prides itself on being incredibly data-driven about meeting consumer needs. But optimizing a Web page to sell the greatest number of hair dryers is a lot different than making a choice about ordering a TV pilot to series.

When any carpetbagging tech company steps outside its core competency and gets into original programming, there’s always the risk of producing bad product that reflects an underestimation of what it takes to manufacture quality

But if the 14 pilots Amazon Studios made available for free on Amazon Instant Video earlier in April are any indication, they’re getting this off on the right foot. Of the eight comedies offered, I was impressed by “Betas,” a tale of young tech entrepreneurs that played like a cross between “Big Bang Theory” and “The Social Network.” Even “Zombieland,” a TV adaptation of the Sony horror pic that doesn’t really sound great on paper, worked well with a cast of unknowns.

It’s not that any of these shows screams instant hit; more to the point, none of them are embarrassing in execution. Rather than appear as if they are the castoffs of broadcast and cable development, which they essentially are, the Amazon slate could stand toe to toe with the content it lost out to at those networks.

That’s important. Like Netflix, Amazon does not want to be seen by either the industry or the marketplace as being a cut below what is being produced elsewhere. It defeats the purpose of getting into original programming.

With its own data-centric orientation and original programming pipeline, Netflix is the obvious comparison with what Amazon is doing. But there’s a pretty stark contrast in their approaches when you consider that Netflix has decided to skip the pilot process entirely and make sizable multi-episode commitments off the bat. “House of Cards” was a big gamble, $100 million for 26 episodes sight unseen. And while it’s premature to call it a success, it certainly wasn’t a failure. Compared with Amazon, those initial investments seem to out-Hollywood Hollywood.

But even if Amazon has developed an uncanny ability to gauge audience tastes, it raises a question: Do viewers even really know what they want? TV history is littered with examples of huge hits that didn’t come into their own, like “Seinfeld” and “Cheers,” until they spent many episodes on the air languishing in the ratings but finding their way creatively in a way that may not be apparent in a pilot. That’s less common nowadays in the shrinking TV ratings universe, though perhaps that very shrinkage becomes a blessing in disguise: Networks might as well be patient if the return on investment is dwindling.

As the broadcast business heads into the thick of its development season, why the networks have yet to try the Amazon route is surprising. What an incredible culture shift it would be if one of the nets threw open the development doors and got audiences involved this early. North Korea would sooner become a democracy.

Filed Under:

Want Entertainment News First? Sign up for Variety Alerts and Newsletters!
Post A Comment 1